Potentiometer debounce?

Thread Starter

Hamlet

Joined Jun 10, 2015
435
I'm replacing a 10turn/100k trimmer with a 10-turn/100k wirewound pot on a smps module, for an adj. bench power supply project.

What I am finding is that my inventory of 100k wirewound potentiometers are all flakey. Normally, I'd just patch one end of the pot to the wiper, & settle things out that way, but with the smps that I'm using, a high ohmic value corresponds to max voltage output on the smps...

When the wiper on the pot slides along the windings, it bounces, loosing continuity, causing voltage spikes.

I was thinking of bypassing one of the legs of the pot (wiper?) to ground with a cap, but none of the three legs of my pot have any
continuity to the ground plane, so I don't know what to do.
 

Papabravo

Joined Feb 24, 2006
15,481
You know that pot is just going to be a boatload of trouble down the line. The customer service time and aggravation is going to destroy you. I'd be thinking of alternatives RSN. That's Real Soon Now. One more thing: hanging a cap on to fix a mechanical problem is just not going to make it.
 

SamR

Joined Mar 19, 2019
3,180
Not a good start. You could try spraying some contact cleaner inside and rotating the pot. Not sure how that would work on a 10 turn pot as they are pretty closed up but I've had some success with standard pots. I'd be looking for a replacement and griping at whoever sold you the flaky pots. Contact cleaner is a patch, not a fix.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
435
I'm not safe for retail, so customer service aint no thing to me. My inventory is for personal enjoyment, except when it leads to frustration. I just want a cheap fix, for my hobby-side interests...
I could source a good pot from an authorized Bourns dealer, but that would cost more than the smps I'm working with, and I wouldn't learn anything.

Come on, switch debounce is a valid concern, why not pots?
 

MrChips

Joined Oct 2, 2009
23,073
A potentiometer has three terminals, yes?
That is so that you can use it as a potential divider.

But suppose you want to use a pot as a variable resistor. You only need to use two of the three terminals, one arm and the wiper.

What do you do with the unused terminal?
Here is good advice. Connect it to the wiper terminal. This will prevent the pot from ever becoming an open circuit.

Edit: sorry, I just read that you have already done that.
 

MrAl

Joined Jun 17, 2014
8,063
I'm replacing a 10turn/100k trimmer with a 10-turn/100k wirewound pot on a smps module, for an adj. bench power supply project.

What I am finding is that my inventory of 100k wirewound potentiometers are all flakey. Normally, I'd just patch one end of the pot to the wiper, & settle things out that way, but with the smps that I'm using, a high ohmic value corresponds to max voltage output on the smps...

When the wiper on the pot slides along the windings, it bounces, loosing continuity, causing voltage spikes.

I was thinking of bypassing one of the legs of the pot (wiper?) to ground with a cap, but none of the three legs of my pot have any
continuity to the ground plane, so I don't know what to do.
What works *sometimes* is to connect a capacitor from each end to the wiper. That's two caps not one. You may need a larger size cap, 1uf, 10uf, maybe even 100uf, depending on the pot value and the circuit values. If you do use polarized caps you have to first find out what the polarity of the wiper is relative to both ends so you know where to connect the cap positive (and negative) leads.

But really any fix depends on the particular circuit topology. To do this right we would need to see the circuit schematic. For example, if the pot functionality can be reversed then we could make it such that when the pot wiper loses contact with the resistance wire inside and goes open circuit, the output voltage goes DOWN. That may be good enough in some cases but not others, so again it is good to know the circuit schematic. The best solution will come from knowing that. The solution with the two caps for example could cause the circuit to become very unstable in some applications.

Also what works sometimes is to place a relatively large value resistance from the wiper to ground so that if the wiper goes open the resistor connects that node of the circuit to ground through a resistance. Again this is only with certain circuits.

This problem is not that unusual. The LM317 type regulators also have this problem. If the pot wiper opens up the entire resistance may suddenly appear in the circuit and cause the output to shoot up.

Another idea is to sense the wiper node voltage level and if it goes out of range cut back the output using another control signal. The wiper can be pulled high or low with a large value resistor and this out of range voltage level would be sensed by the over range circuit when the wiper goes open.

Another fail safe idea is to have a secondary means of regulation. If the pot wiper opens the secondary regulation circuit kicks in and regulates the output to maybe 0.5 to 1.0 volts higher than the primary setting allowed and sets off an alarm which can be just an LED or a buzzer.
This secondary regulation setting has to be set independently of the primary setting so it means having a second pot to adjust, at a time different than the time the primary pot is being adjusted.

Depending on the circuit, you may even be able to use just two pots and maybe some diodes. You have to adjust each pot in turn to get the output right, but there is always a pot that is not being adjusted ready to take over in case the other one opens up.

Just some ideas, but again seeing the circuit is the best way to go about this because the wrong solution could cause more problems than it solves.
 

Tonyr1084

Joined Sep 24, 2015
5,810
switch debounce is a valid concern, why not pots?
Because switch contacts bounce against each other when they close. Pots don't bounce, they slide. Even wire wound pots slide. And they should contact more than just one wire at a time. Cheap pots are just that - cheap. Undependable. A 10 turn wire wound pot wiper will move quite slowly as it crosses from one wire to the next. At transition from one wire to the next it should make contact with both at the same time. The steps in resistance value should be almost nothing.

However, pots do become dirty. "Stuff" can build up on the surface making the wiper lose contact, causing a scratchy sound in audio equipment. Contact cleaner can help. Years ago and early in my electronics journey I was asked to fix a stereo for a girlfriend. When you turned up the volume (or down) there was terrible scratching sounds. I cleaned it with Naphthaline (lighter fluid). Of course it evaporated before adjusting volume, and it helped. But its benefit lasted only a few days before it was back to being scratchy. The best answer is to replace the pot with something reliable. Non-polarized caps across terminals 1 & 2 and 2 & 3 will help with the issue; but there's no substitute for quality.
 

Papabravo

Joined Feb 24, 2006
15,481
I'm not safe for retail, so customer service aint no thing to me. My inventory is for personal enjoyment, except when it leads to frustration. I just want a cheap fix, for my hobby-side interests...
I could source a good pot from an authorized Bourns dealer, but that would cost more than the smps I'm working with, and I wouldn't learn anything.

Come on, switch debounce is a valid concern, why not pots?
Because pots are not designed to be make or break switch contacts. If they are doing that then they are defective period. You can try whatever you want to attempt a fix, but ultimately you can't will physics to work the way you wish it would. When I was a young engineer one of the grizzled veterans said: "You know, everything you're working on today will be in the landfill ten years from now". I'm with the school of thought that says: "bin them and move on".
 

Thread Starter

Hamlet

Joined Jun 10, 2015
435
Sometimes the practical solution is to throw out defective components. It is also cathargic.
Yes, I break things open after they burn, inspect, and into the scrap-can I keep at my feet. These pots will labled as sh/t, & go into my nichrome wire bin.

Maybe I can workout a fine/couse adj. scheme with single-turn pots. Thank you all for the input.
 

SamR

Joined Mar 19, 2019
3,180
Maybe I can workout a fine/couse adj. scheme with single-turn pots.
That thought did cross my mind as it is a valid solution to stepping up from the small Bourne type trimmers when greater wattage is called for and a bigger multiturn not available.
 

MrChips

Joined Oct 2, 2009
23,073
As already stated, the work-around solution will depend on the circuit and its application.
Decide what you want to happen if the wiper loses contact.
If this is for setting the output voltage of a PSU then you would assume that you want the output voltage to fall and not rise. Install a resistor from the wiper to a node that will cause the output voltage to fall (or temporarily move towards a desired voltage). You can use two resistors in series along with a smoothing capacitor so that there is some acceptable time constant.

Edit:

Pot debounce.jpg

VR1 and CONTROL VOLTAGE is your existing circuit.

R1, R2, C1 are connected to any SAFE VOLTAGE which could be GND or any voltage that temporarily moves the CONTROL VOLTAGE to a safe control value depending on the application. I would have to see the actual circuit in order to give specific recommendations. You might be able to omit R2 and SAFE VOLTAGE altogether.
 

Thread Starter

Hamlet

Joined Jun 10, 2015
435
As already stated, the work-around solution will depend on the circuit and its application.
Decide what you want to happen if the wiper loses contact.
If this is for setting the output voltage of a PSU then you would assume that you want the output voltage to fall and not rise. Install a resistor from the wiper to a node that will cause the output voltage to fall (or temporarily move towards a desired voltage). You can use two resistors in series with along with a smoothing capacitor so that there is some acceptable time constant. If you do not understand I will post a schematic.
I'm all ears. I'm thinking this would be good practice no matter the quality of the pots.
 

MrChips

Joined Oct 2, 2009
23,073
On second thoughts, I think this is a better solution.

The voltage on C1 will always settle to the desired control voltage.
R1 is added so as not to slow down the response when the pot is adjusted manually.

If the control pot is an essential part of the feed-back control loop you will have to consider the effect of introducing any extra delay into the loop.

Pot debounce2.jpg
 

Thread Starter

Hamlet

Joined Jun 10, 2015
435
On second thoughts, I think this is a better solution.

The voltage on C1 will always settle to the desired control voltage.
R1 is added so as not to slow down the response when the pot is adjusted manually.

If the control pot is an essential part of the feed-back control loop you will have to consider the effect of introducing any extra delay into the loop.

View attachment 232178
Okay, a 1k and 10uF to ground. Now my output is pulsating dc, swinging perhaps 10v, but I no more wild swings of over-voltage... and responsiveness of pot seems good.
 

MrChips

Joined Oct 2, 2009
23,073
Okay, a 1k and 10uF to ground. Now my output is pulsating dc, swinging perhaps 10v, but I no more wild swings of over-voltage... and responsiveness of pot seems good.
Is the output supposed to pulsate?
If not, then keep increasing the resistor value until it stops pulsating.

Edit: Wait a minute. Can you please post your circuit diagram?
 
Top